Skill Prompt Generator
A set of skills for Claude Code and Codex that assemble detailed image prompts from an element library and composition schemes
- 1,525 ★ on the repository
- Last commit
Extended access
Why this level
- The skills run a local Python engine and read the project's element database
Install
- Works in
Manual install
git clone https://github.com/huangserva/skill-prompt-generator.git
cd skill-prompt-generator
pip install -r requirements.txtAfter cloning, the skills under .claude/skills/ are picked up by Claude Code automatically. Codex uses the .codex/ directory.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skills run a local Python engine and read the project's element database
This is third-party code. Review the repository files before installing.
What it does
The set gives the agent twelve skills that turn a short natural-language request into a full image-generation prompt. The agent detects the domain on its own, portrait, art, design, product photography or video, and calls the right specialist. A prompt is assembled from an element library and a corpus of community prompts, with a consistency check and variable sampling on top so results do not repeat. There are three modes: a single-domain portrait, a cross-domain scene that automatically combines several domains, and poster and card design with palette, border and decoration sets. A local Python engine runs under the skills: it stores elements in a database, while the agent fills in missing semantic detail such as specific characters or techniques from its own knowledge. The text and documentation are mostly Chinese, and the resulting prompts come out in English.
Who it is for. For designers, content makers and marketers who need detailed, repeatable prompts for AI image generation.
Good fit when
- You need a full prompt for a portrait, art piece, product shot or poster
- You need a complex scene where elements from several domains should combine
- You want variety in prompts instead of repetitive phrasing
Not a fit when
- You need the image itself, not prompt text: the set does not render images
- The Chinese documentation is a barrier and you do not want to dig into the skill structure
Example request
Assemble a prompt for a cinematic East Asian portrait with dramatic lightingLimitations
The set produces only prompt text; the image itself is rendered by a separate model. A Python engine runs under the skills, so Python 3.8 or newer and dependencies from requirements.txt are required. The element library does not cover every possible IP or character, and the agent fills gaps from its own knowledge. The documentation and interface are mostly Chinese. The repository states no license.
How to disable. Remove the skills directory from the project's .claude/skills/ or from ~/.claude/skills/, and optionally delete the cloned repository folder.
FAQ
Does the set draw images?
No. It assembles a detailed prompt text; the generation is done by an image model of your choice using that prompt.
Do I need Python?
Yes. A local Python engine runs under the skills; you need Python 3.8 or newer and the dependencies from requirements.txt.
Discussion
No questions or reviews yet.
Related
SkillsA Claude Code skill that compiles a vague request into a finished gpt-image-2 prompt and checks it with a validator script
Five image-generation skills: Qwen-Image prompts, logos with shared geometry, diary comics, transparent PNGs and process dioramas
Claude Code subagent that turns a rough image idea into a detailed prompt using a seven-part formula for image generators.
image-prompt-reverse analyzes an uploaded picture and writes a generator prompt together with a negative-term list.
A skill that searches a library of 10,000+ image prompts, shows sample images and tailors the chosen prompt to your article or script
The skill turns paper text, methods and captions into structured prompts for scientific figures, tuned to the chosen image model.
README in short
The README describes a system of twelve skills for Claude Code and the Codex CLI that generates image prompts from a 1246-element library and a corpus of community prompts. The user works through the skills rather than Python directly: the agent detects the domain and calls the right specialist. Version 2.0 adds cross-domain queries, a design-system integration with palette sets and three generation modes: portrait, cross-domain and design. Installation is git clone plus a pip install of dependencies; skills under .claude/skills are picked up automatically, and individual ones can be added via npx skills. The documentation is Chinese and the prompts come out in English.
README badge
Original SKILL.md
--- name: intelligent-prompt-generator description: intelligent image prompt generator v2.0, supports portrait, cross-domain and design modes, with semantic understanding, common-sense reasoning and consistency checks --- # Intelligent Prompt Generator Skill v2.0 You are an intelligent prompt-generation expert with semantic understanding, common-sense reasoning and consistency checks. The system supports three generation modes: Portrait for pure portrait photography using the portrait domain, Cross-Domain for complex scenes that combine several domains automatically, and Design for posters and cards using SQLite elements plus YAML variables for palettes, borders and decorations. The system detects the request type and picks the best mode: a subject without complex needs maps to portrait, a subject with action or effects maps to cross-domain, and design-style keywords map to design.